Build a serverless ML algorhytms for user engagement predictions

We need to build a solution that will calculate:

- The probability of a user opening a message over the next 7 days
- The probability of a user clicking a link within a message over the next 7 days
- The probability of a user opting out of messaging over the next 7 days
- The expected page views a user will generate over the next 30 days
- The expected order value if a purchase occurs in the next week
- The probability of a user making a purchase over the next 1, 7, or 30 days
- The total revenue from all predicted purchases over the next 30 or 365 days

The solution should utilize AWS serverless services, so we would not need to manage servers.

The solution should be multi-tennant as we have several thousands of customers that would use this.

Example data set will be provided.

IMPORTANT:
- you need to describe the tech stack / architecture in your proposal, or you wont be accepted
- you need to describe the costs of running the solution, or for eg. 10000 predictions